Experiment with Reinforcement learning 0

Описание к видео Experiment with Reinforcement learning 0

The planar arm's pen must reach the red circle with a line as straight as it can.
Their are 8 possible actions to control the arm's, each servos is either going one unit forward, one unit backwards or stationary. The environment use the distance to the target and the angular deviation from the ideal path to calculate the rewards.
It is programmed in python using R. Sutton's RLToolkit

Комментарии

Информация по комментариям в разработке

Похожие видео